*4.2. Spatial Pattern Analysis of Potential Land-Use Conflict*

Table 3 shows that the *Moran's I* values for land use conflicts in Lin'an District in 2008, 2013 and 2018 were 0.238, 0.253 and 0.232, respectively, indicating that the comprehensive index of land-use conflict in Lin'an District showed significant and positive global spatial autocorrelation.

To reveal the spatial relationships among local spatial units in Lin'an District, this research further calculated the local spatial autocorrelation index *LISA* and obtained clustering and significance results for land-use conflict. In the quantitative clustering results (Figure 3), except for the non-significant spatial units, the spatial units were dominated by high-high aggregation, which accounted for 5.54%, 5.92% and 5.04% of the units in 2008, 2013 and 2018, respectively. The spatial units with low-low aggregation changed significantly, increasing from 4.59% in 2008 to 5.70% in 2013 and then decreasing to 4.65% in 2018. Over the 10 years, the proportion of spatial units with low-high aggregation remained between 4.93% and 5.67%. The proportions of spatial units with the above three clustering grades were similar; all of them were approximately 5% and increased first and then decreased. The proportion of spatial units with high-low aggregation was the smallest, but it continued to increase, reaching its highest proportion of 0.81% by 2018. The proportion of nonsignificant spatial units was the highest; it remained higher than 80% during the period 2008–2018 and showed a trend of first decreasing and then increasing. The above results show that the spatial unit clustering of land-use conflict was the most significant in 2013.

**Figure 2.** The degrees of land-use conflict between agricultural and ecological spaces in Lin'an District, 2008–2018. (**a**) 2008. (**b**) 2013. (**c**) 2018.


**Table 3.** Changes in the global spatial autocorrelation of land-use conflict in Lin'an District, 2008–2018.

The distribution areas of high-high aggregation, low-low aggregation and low-high aggregation spatial units in the clustering results were relatively similar (Figure 4). The high-high aggregation spatial units were mainly distributed zonally. In addition, there were obvious aggregation areas in the eastern region. The cultivated land, garden land and water areas were interlaced in these spatial units, and the land-use types were more complex. The low-low aggregation spatial units were scattered, and the streets of each town had low-low aggregation spatial units. Their distribution in Sun town was the most obvious; this area was mainly grassland, the surrounding land-use type was uniform, and the conflict level was low. The low-high and high-low aggregation spatial units were distributed on the periphery of the high-high and low-low aggregation spatial units, respectively, and the latter had a smaller area of distribution. The non-significant spatial units were the most widely distributed and had a continuous distribution; these units were mainly forests.

**Figure 3.** Changes in the local spatial autocorrelation of land-use conflict in Lin'an District, 2008–2018. (**a**) Clustering results. (**b**) Significant results.

**Figure 4.** Local spatial autocorrelation of potential land-use conflict in Lin'an District, 2008–2018. (**a**) 2008. (**b**) 2013. (**c**) 2018.

The spatial units with *p* = 0.001 in the significance results (Figure 5) have obvious clustering areas, mainly in the eastern part, Qianchuan town, Heqiao town, Longgang town and Qingliangfeng town. The spatial units with *p* = 0.01 and *p* = 0.05 were distributed sequentially at their periphery, with the latter having the largest distribution area except for the non-significant spatial units. The distribution characteristics of the non-significant spatial units were consistent with the clustering results.

**Figure 5.** Significant results for local spatial autocorrelation of potential land-use conflict in Lin'an District. (**a**) 2008. (**b**) 2013. (**c**) 2018.
